This disclosure relates generally to video analysis techniques for video editing, video navigation and video summarization, and in particular to improve video analysis techniques based on the determination, generation and analysis of image metadata for prioritization, classification and summarization of video data.
Captured video may contain redundant, unnecessary and/or poorly filmed sections of video. Such undesirable forms of video may be caused by any number of factors. For example, a shaky camera, poor exposure, redundant scenes, partially obstructed scenes and unintended scenes that were not intended to be captured can exist in captured video.
Although various manual techniques may be known for lessening individual ones of some such characteristics, such techniques are often provided as part of a cumbersome or unwieldy process. For example, such techniques are often applied one-at-a-time and without any consideration of dependencies among other techniques that will also be applied or that have been applied. Such a process can be time consuming and require considerable knowledge of the particular techniques.